README.md

In lizhongliu1996/PredictMisc: package that use three methods to predict methylation regions

PredictMisc

package that do prediction analysis for methylation regions

Required Packages

Because you must install this package from GitLab, you will need to have a few packages installed on your machine before you can build this package.

• Aclust: install this from https://github.com/tamartsi/Aclust. This package also suggests the Bioconductor package AnnotationDbi.
• caret
• data.table
• dplyr
• pathwayPCA: install this from https://github.com/gabrielodom/pathwayPCA
• pROC
• tibble

purpose

Typically, when using doing methylation predictions, there are 450k number of CpGs, how to select predictors is a tricky problem. Usually, people choose CpGs that are most significant or CpGs with most varilibity, here we present a new method using differential methylation regions to select CpGs as prodictors

lizhongliu1996/PredictMisc documentation built on Aug. 23, 2019, 5:55 a.m.